Gigabyte's 1080 Waterforce has thermal pads everywhere it should.
 
I just got a 1060 3GB from EVGA though for my HTPC, and I'll probably be applying some thermal pads anyway just as well.

i use to have a 295x2 (it was 2 290's on a single water cooled card). it would run pretty warm, which makes sense since it was 2 gpu's on a single 120mm rad. anways, i replaced the thermal compound with grizzly kryonaut, replaced the vram pads with grizzly minus 8 pads, and replaced/added some fujypoily sarcon pads (expensive as F) on the other chips next to the gpu (cant remember what they are called).
 
so before, i believe it was hitting high 70's, maybe even 80's at stock. after the "upgrade" it was down to the high 60's very low 70's with an OC on the card.
 
my point, replacing thermal compound and replacing/adding pads made more of a difference then i would have expected. it cost me like $90 for everything, but it was worth it, and i at least still have plenty of the thermal paste left over for future cards/cpu installs.

OK, that mean problem practically not exist or in worse case is easy to resolve with thermal pads.
Without doubt GTX1080 FTW Hybrid is best choice for average gamer.
There is only one problem, card is expensive and I hope NVIDIA will give us GTX1080Ti for same price.
Than GTX1080Ti FTW Hybrid is acceptable for same price and we could wait second generation of Volta.
Or there are other option, if price of GTX1080 FTW Hybrid drop on 600 euro example after second generation show up.
Situation can't be clear or better before AMD show up. Last 2 years NVIDIA fans wait Radeon more than their buyers.
With one wish, to price of GeForce drop and than people could make better upgrade.

Iluv2raceit

jlp209

Iluv2raceit
 
What I find disturbing about EVGA's response that the GTX1080 FTW Hybrid is not affected.  I checked myself using a high intensity LED flash light and confirmed that the my FTW Hybrid card does NOT have a thermal pad in place that covers the backside of the VRMs and makes contact with the back plate.  Nothing but air space between, which is bad.  I have already ordered thermal pads and will put them in place to ensure adequate cooling of the VRMs.   Better safe than sorry is my motto.  Still, it pisses me off a bit that EVGA really feels they have to cut corners on thermal pads that cost only a $1 or so.  The assumption that the use of a Hybrid cooler will negate the VRMs from overheating is short sightedness at best.  And interestingly enough, the EKWB back plate for my water cooled Titan XP does have a thermal pad that is provided to transfer the heat from the VRMs to the back plate.  So, why doesn't EVGA offer the same solution for the FTW Hybrid cards as well??

Were you able to order the thermal pads from EVGA or did you get generic ones that will fit? My latest card / 3rd FTW Hybrid (also had a Classified) has been the most stable by far. I am very curious why it is so different. Prior cards, aside from black screen + fans 100%, would always hit power + voltage limit at same time causing massive downclocking, while temps stayed under 65 degrees. At stock I'd boost to 2025 mhz and still got downclocked to 2012 mhz and 2000 mhz for no reason that made sense to me (on a Hybrid with temps under 60 degrees). Maybe it was related to inadequate VRM cooling, I'll never know.
 
The Classified I had hit the temp limit in Precision (also at different times hit voltage and power limit), yet the core never went over 65 degrees. Maybe that had something to do with the VRM issue. Bizarre behavior, returned it to Microcenter. 
 
My current FTW Hybrid stays at 2113 mhz regardless of game or benchmark almost all the time, with the occasional dip down to 2101 mhz. But even when that happens, I've never hit voltage, temp, or power limit. A down clock of 12 mhz / 1 bin is normal and what my prior cards should've been doing. So confused by my experiences so far, definitely the most issues I've ever had with any GPU ever. EVGA support has been great though, luckily. 

The issue has nothing to do with GPU temps and is directly associated with the VRM temperatures, which are not monitored by the card's BIOS or any other tool for that matter since there are no thermostats near the VRMs.  This is something altogether a different problem that is fortunately easily fixed by just applying some thermal pads to the back of the PCB so that the heat can be dissipated through the back plate.  EVGA would not allow for me to submit the request for the thermal pads, so I had to order them from Amazon.

Did this fix your VRM problems by chance? I just RMA'd 1 of the 2 Hybrids I have because I kept crashing due to black screen/100% fan issue that people are complaining about. EVGA really only confirmed issue with the FTW cards and not the Hybrids saying the issue that people are having shouldn't be affected by it. And most threads are just about the FTW cards. 
 
If it's as simple as adding thermal pads to the VRM between the board and the back plane, then I much rather go that route then waiting weeks for the RMA to come, and submitting another RMA for the 2nd card. Also, I'm guessing there's nothing that can be done about the inductors like the installation for the FTW cards say.

so in other words should hybrid owners be worried about the vrm heating issues?

silixx
so in other words should hybrid owners be worried about the vrm heating issues?

I'm now also wondering about this. I actually started a thread about this question this afternoon and was told the Hybrid cards weren't affected. However, reading this thread is giving me seconds doubts.
 

 

It's easy. 
 
If the VRMs are affected by overheating issues, the hybrid cards are not affected. The active cooling is deemed adequate. Contrary to what earlier posters mentioned, the fan ON the card is completely meant to cool the VRM and is in direct contact via the thermal pads. There is no need for thermal pads on the backside of the PCB as the backplate is not needed for heat dissipation. This is different from the ACX 3.0 model, where both the mid- and the backplate are meant as surface increase for passive heat dissipation of VRM and VRAM.
 
If the VRMs are affected by QC issues from the supplier, the hybrid cards are affected. According to EVGA this QC-issue has been resolved back in august. Cards manufactured and shipped after 8/31 should not be affected. Cards manufactured and shipped before might bear a 4% chance to be faulty. Warranty will cover it in case of faulty VRM. The chance that your card might "explode" is actually really low. IF you are affected by a faulty VRM it is more likely to encounter the blackscreen + 100% fan issue.

Angier_1985
It's easy. 
 
If the VRMs are affected by overheating issues, the hybrid cards are not affected. The active cooling is deemed adequate. Contrary to what earlier posters mentioned, the fan ON the card is completely meant to cool the VRM and is in direct contact via the thermal pads. There is no need for thermal pads on the backside of the PCB as the backplate is not needed for heat dissipation. This is different from the ACX 3.0 model, where both the mid- and the backplate are meant as surface increase for passive heat dissipation of VRM and VRAM.
 
If the VRMs are affected by QC issues from the supplier, the hybrid cards are affected. According to EVGA this QC-issue has been resolved back in august. Cards manufactured and shipped after 8/31 should not be affected. Cards manufactured and shipped before might bear a 4% chance to be faulty. Warranty will cover it in case of faulty VRM. The chance that your card might "explode" is actually really low. IF you are affected by a faulty VRM it is more likely to encounter the blackscreen + 100% fan issue.

Awesome, your post makes sense :) I think I know enough now to order my 1070 hybrid. 
 
Another thing I'm really wondering about. I'd like my upcoming build to be as silent as possible, but I'm working in a really small case (Ncase M1) where heat is an issue. The thing I like about all air cooled GPU's is the 0db mode.
 
It has been quite a while for me building a PC and how fan control on the motherboard works. But is it possible to connect the fan of the radiator to the motherboard (3-pin, cause the only 4-pin is being used by my CPU cooler) and create a 'fan curve' based on the temp of the GPU? Meaning I can turn of the fan, or lower the RPM significantly, in idle to sort of mimic the 0db mode of the air cooled cards? 

There are tools that can control such a fan via the GPU temps, like speedfan. Most often, it is more advisable to look for the sweet spot of sound and heat dissipation and have your fan run at this setting all the time. A sudden change in noise is more noticable than a constant humming as background noise and thus even a fan that features no idle mode can still appear "silent" as long as it is quieter than for example your average volume setting or other background noise in your room. I usually aim to be quieter than any A/C.

Angier_1985
There are tools that can control such a fan via the GPU temps, like speedfan. Most often, it is more advisable to look for the sweet spot of sound and heat dissipation and have your fan run at this setting all the time. A sudden change in noise is more noticable than a constant humming as background noise and thus even a fan that features no idle mode can still appear "silent" as long as it is quieter than for example your average volume setting or other background noise in your room. I usually aim to be quieter than any A/C.

Good to know I can use software like speedfan to control the fan of the hybrid card. I'll make sure to just use a really low RPM instead of turning off the fan entirely.